Research has shown that the anger management styles of both anger-in (suppression of anger) and anger-out (direct verbal or physical expression of anger) may be associated with elevated chronic pain intensity. Only the effects of anger-out appear to be mediated by increased physiological stress responsiveness. Given the catecholamine-sensitive nature of pain mechanisms in complex regional pain syndrome (CRPS), it was hypothesized that anger-out, but not anger-in, would demonstrate a stronger relationship with chronic pain intensity in CRPS patients than in non-CRPS chronic pain patients. Thirty-four chronic pain patients meeting IASP criteria for CRPS and 50 non-CRPS (predominantely myofascial) limb pain patients completed the McGill Pain Questionnaire-Short Form (MPQ), the Anger Expression Inventory (AEI), and the Beck Depression Inventory (BDI). Analyses revealed no diagnostic group differences in mean scores on the anger-in (AIS) and anger-out (AOS) subscales of the AEI, or on the BDI (values of P>0.10). Results of general linear model analyses revealed significant AOS x diagnostic group interactions on both the sensory (MPQ-S) and affective (MPQ-A) subscales of the MPQ (values of P<0.05). In both cases, higher AOS scores were associated with more intense chronic pain in the CRPS group, but with less intense pain in the non-CRPS limb pain group. Inclusion of BDI scores as a covariate did not substantially alter the AOS x diagnostic group interactions, indicating that these AOS interactions were not due solely to overlap with negative affect. Although higher AIS scores were associated with elevated MPQ-A pain intensity as a main effect (P<0.05), no significant AIS x diagnostic group interactions were detected (values of P>0.10). The AIS main effect on MPQ-A ratings was accounted for entirely by overlap with negative affect. Results are consistent with a greater negative impact of anger-out on chronic pain intensity in conditions reflecting catecholamine-sensitive pain mechanisms, presumably due to the association between anger-out and elevated physiological stress responsiveness. These results further support previous suggestions that anger-in and anger-out may affect pain through different mechanisms.